The Social Motivation Hypothesis for Prosocial Behavior

(Forthcoming in Philosophy of the Social Sciences)

Marion Godman (1,2), Michiru Nagatsu (1) and Mikko Salmela (1)1

(1) University of Helsinki, (2) University of Cambridge

Abstract

Existing economic models of prosociality have been rather

silent in terms of proximate psychological mechanisms. We

nevertheless identify the psychologically most informed

accounts and offer a critical discussion of their hypotheses

for the proximate psychological explanations. Based on

convergent evidence from several fields of research we argue

that there nevertheless is a more plausible alternative

proximate account available: the social motivation hypothesis.

The hypothesis represents a more basic explanation of the

appeal of prosocial behavior, which is in terms of anticipated

social rewards. We also argue in favour of our own social

motivation hypothesis over Robert Sugden’s fellow-feeling

1 These authors contributed equally to the work and are listed in alphabetical order. Different versions of this paper were presented at the following interdisciplinary conferences: British Society for the Philosophy of Science (Exeter), Joint Action Meeting (Berlin), European Network for Social Ontology (Helsinki), European Network for Philosophy of Social Sciences (Venice), and also at Philosophy of Science seminar inHelsinki. We thank all the audiences for their valuable comments and suggestions. We would also like to thank three anonymous reviewers for this journal whose comments have led to a much improved paper. Marion also wishes to thank the Volkswagen Foundation whose European Platform for Life Sciences, Mind Sciences, and the Humanities has supported her involvement in the interdisciplinary project on social interaction (http://social-interaction.eu/). She has also learnt a lot from co-participants in the project: Vivian Bohl, Mog Stapleton, Christoph Teufel, Wouter van den Bos and Marijn van Wingerden. Michiru wants to thank Chiara Lisciandra for drawing attention to Clavien & Klein (2010) and Lisciandra et al. (2013).

1

account (due originally to Adam Smith). We suggest that the

social motivation hypothesis not only stands as a proximate

account in its own right; it also provides a plausible scaffold

for other more sophisticated motivations (e.g. fellow-

feelings). We conclude by discussing some possible implications

the social motivation hypothesis has on existing modeling

practice.

Table of contents

1. Introduction

2. Economic models of prosocial behavior

2.1 Social norms approach

2.2 Team reasoning approach

2.3. Summary

3. The “fellow-feeling” hypothesis

4. The social motivation hypothesis

5. Social motivations as scaffolders

6. Conclusion

1. Introduction

It is widely accepted that a distinctive aspect of human

sociality is the pervasive engagement in “prosocial behavior”,

which can be roughly characterized here as an individual’s

choice to forgo her immediate material interests in order to

benefit other(s) or a group to which she belongs. Many

behavioral economic models hypothesize some proximate

motivations in addition to purely material self-interest in

order to explain such prosocial behavior. However, the proposed

psychological mechanisms are often rather casually proposed to

accommodate behavioral data in stock experimental games, and

2

are often poorly informed by psychological research. Moreover,

contrary to the stated ambitions, this strategy so far remains

unsuccessful in delivering a unified picture of what motivates

prosocial behavior (Clavien and Klein 2010).

This paper was triggered by the following observation: perhaps

there are too many models and not enough attention to the

empirical work in various areas of psychology. It seems as

though we should try to get the psychological (proximate)

mechanisms right before modelling a particular motivation as an

augment of an individual utility function, or presuming that

such modelling strategy is indeed the best way to proceed. We

support this move while acknowledging that like any empirical

hypothesis, psychological ones are fallible and in principle

open to revision. Still, we hope that this paper will not only

provide the reader with a psychologically realistic and

plausible proximate account of prosocial behavior, but also

prompt some general methodological reflections on how

psychological research bears on existing modelling practice.

We proceed as follows. In section 2 we consider the two

psychologically most informed economic modelling strategies for

explaining prosociality, and argue that they are, at least as

stands, unsatisfactory with respect to providing proximate

accounts of prosocial behavior. In section 3, we consider Adam

Smith’s fellow-feeling hypothesis revived by Robert Sugden

(2002), as a more promising alternative for the motivations for

norm following/team reasoning. In section 4, we present a

different social motivation hypothesis as well as some

convergent evidence for it and in section 5, we argue that

social motivations might be understood as a scaffold for the

3

more specific rewards involved in fellow-feelings and,

analogously, as a general scaffolder for other candidate

proximate accounts of prosocial behavior. Section 6 concludes

with some methodological reflections.

2. Economic models of prosocial behavior

Prosocial behavior, such as cooperation in the prisoner’s

dilemma (PD) and public goods games (PG), fair offers in

ultimatum bargaining games (UG), and trusting and sharing in

trust games, is a fairly robust experimental phenomenon -- even

in those experiments where one is careful to ensure that the

subjects believe that interactions are anonymous and that there

is no future interaction with other subjects (see e.g. Camerer

2003, ch. 2 for a review).

We can identify at least four approaches within the rational

choice tradition that purport to explain such experimental

results as well as our prosocial behavior in analogous real

life situations. The first two are (i) the rational cooperation

in repeated games approach, according to which people

mistakenly behave prosocially in one-shot games because humans

have evolved in repeated interactions where cooperation is

rational (e.g. Binmore 2006); and (ii) the social preference

approach, which models people’s motivations as stable “social

preferences” for equity, fairness, etc. that are defined in

terms of the material payoffs of games (e.g. Fehr and Schmidt

1999). Although popular among behavioral economists, there is

convincing empirical evidence against these hypotheses (see

e.g. Gächter and Thöni 2007 for a review). Moreover, these

accounts do not explicitly address motivations of prosocial

behavior: the first (i) simply assumes such motivations to be

4

explanatorily irrelevant; and the second (ii) crudely equates

prosocial motivations with preferences for different

distribution of players’ material payoffs in games. We

therefore focus on the two other approaches in this paper: (1)

social norms; and (2) team reasoning.

Both (1) social norms and (2) team reasoning seem to hold the

most promise for explaining a range of experimental results

(see Tan and Zizzo (2008); Guala et al. (2013) for empirical

evaluations). The approaches are also (though not always)

accompanied by explicit discussions of what motivates norm

adherence and team reasoning, respectively. However, after a

discussion of each approach, we will briefly argue that these

proximate accounts are ultimately unsatisfactory.

2.1 Social norms approach

The social norms approach has been gaining popularity due to the

growing evidence that specific expectations (beliefs) as well

as preferences influence prosocial behavior in economic games.

Unlike the social preferences approach, which assumes some

stable other-regarding preferences at work in all contexts, the

social norm approach adopts the idea of conditional preference.

One of the main proponents of the approach, Christina Bicchieri

thus argues that the preference to follow a norm or behavioral

rule is activated only if the individual believes (a) that

enough others will follow the rule (“empirical expectation”)

and (b) that enough others will expect her to follow

(“normative expectation”) (Bicchieri 2006). Due to this

conditionalization of preferences as relative to certain

expectations and perhaps also what norms are identified, the

5

social norms have a positive explanation for why we should

expect (some) prosocial behavior to occur in some contexts

(e.g. ultimatum games with humans) and not in others (e.g.

ultimatum games with computers) (see Bicchieri 2006, p. 125,

also pp. 168-170). Once these epistemic conditions of norm

identification are met, Bicchieri’s model predicts that

players’ payoffs in a game are transformed. That is, players’

payoffs for different outcomes of the game are subjectively

modified to take into account players’ taste for certain social

norms (of fairness, cooperation, etc.). This changes the payoff

structure of the Prisoner’s Dilemma and other economic games

by, for example, rendering defection or “unfair” behaviour less

attractive, which in turn makes their overall self-interest

aligned with prosocial norms. Social norms “tug at our utility”

of what is economically rational (Paternotte and Grose 2012).

The approach thus models individual utilities such that they

involve preferences for norm conformity.

But what precisely motivates people to follow social norms,

once the epistemic conditions of norm identification are met?

Insofar as the social norm approach can go beyond just

describing the behavior itself in terms of conditional preferences

or stipulate that the mere identification of the norm itself

automatically triggers norm following, the model owes us a

proximate explanation for why agents are motivated to follow the norm and

adhere to others’ expectations.2 In other words, the mere activation of

a social norm (or an expectation that there is such a norm at

2 Although Bicchieri explicitly states (2006, p. 3) that her belief-desire model of choice is compatible with the possibility that our norm following behaviour occurs subconsciously without deliberate decision, she does not understand this folk psychological model as an “as if” model.

6

work) is not sufficient since there must be something that

explains why empirical and normative expectations matter to

individuals.

The same question is provoked by Herbert Gintis’s (2010) model

of social norms, which, just like Bicchieri’s, presupposes that

people have preferences to follow social norms.3 Gintis (2011)

however stresses that certain moral values and character virtues lead

individuals to voluntarily conform to norms. These values and

virtues include promise-keeping, fairness, and honesty, and

they manifest as behavioral regularities that facilitate

cooperation and enhance social efficiency (Gintis 2011, 884).

According to Gintis, individuals internalize these behavioral

regularities through processes of socialization where the moral

values and character are instilled by the interplay of affect

and authority and induce us to comply and conform to the

specified norms that promote prosocial behavior. This

internalization means agents behave prosocially even if there

is no possibility of being penalized for selfish behavior and

being rewarded for prosocial behavior. “Morality, in this way,

is doing the right thing even if no one is looking” (Gintis

2010, 252).

Now, Gintis admits that moral character and virtues do not

always guarantee norm abidance as other conflicting interests

are also present. Yet insofar as we act prosocially, Gintis

believes that the central motivation for such behavior is

normative or moral: doing the right thing in the situation.

3 Gintis (2010) models social norms as a coordination device (the “choreographer”) that enables players to choose one from multiple Nash equilibria, using the concept of correlated equilibrium. See for furthercomparisons of Biccheri’s and Gintis’s models.

7

Still, it seems that Gintis might be too optimistic in thinking

that moral and character virtues alone are responsible for

inducing norm abidance. Recent empirical work in moral

psychology shows that those moral principles and character,

including virtues, that are used to justify actions, do not

appear to have any ‘formative’ role in our decision-making (for

a review, see Merritt et al. 2010). Some even suggest that

moral reasoning is a result of confabulation (Wheatley & Haidt

2005; Schnall et al 2008). Either way, the role of moral traits

and virtues as proximate explanations has not been confirmed to

the extent assumed by Gintis.4

Bicchieri (2006, 13-25) provides a more nuanced and extended

proximate account. She suggests three types of motivations for

norm adherence: (i) fear of resentment (and of more

consequential punishment), (ii) desire to please others’

expectations and preferences, and (iii) belief that others’

normative expectation is legitimate. Bicchieri places most

credence in (iii) and argues that acceptance of certain norms

(e.g. cooperation in PG or fair offer in UG) as reasonable or

legitimate, is the more important proximate explanation. In

interactions with family and friends, a generic desire to meet

others’ expectations will suffice, but in most other

interactions with socially distant others she contends that

generic desires (i) and (ii) have little motivating power, and

so the beliefs about social norms being legitimate are

required.

4 In addition, Gintis does not specify the epistemic conditions under which once-internalised norms are triggered (see Paternotte and Grose 2012).

8

It is undeniable that people are sometimes motivated to meet

others’ expectations because they find those expectations

reasonable or legitimate. But it seems norm conformity is not

always associated with the perception of reasonableness. For

example, Solomon Asch’s (1955) classic experiments show that an

individual’s easy judgment about the length of a line was

significantly influenced by the answer given by other members

of a group when the answer was obviously wrong but unanimous.

In such cases, conformity alone seems to motivate even if the

expectations of others are unreasonable. Thus conformity seems

to be de-coupled from the perceived reasonableness of the norms

(see Lisciandra et al. 2013 for a similar conformity effect on

normative judgements). In other words, people seem to behave as

if they just want to conform to the majority’s opinions and

preferences, regardless of whether it is reasonable to do so or

not.

In fact, once we turn to Bicchieri’s own model, motivations of

types (i) and (ii), seem required in order to derive (iii). In

her account of conditional preferences, Bicchieri distinguishes

two types of “normative expectation”, i.e. one’s belief that a

sufficiently large subset of population expects one to conform

to a behavioral regularity in a type of situations. The first

type is an empirical expectation (i.e. belief that others think one

will conform), and the second normative (i.e. belief that others

think one ought to conform) (2006, 15).5 What motivates us to act

according to such expectations? It seems that viewing such

expectations as legitimate or reasonable is not going to do the

5 So Bicchieri’s terminology is slightly unfortunate. When she calls two epistemic conditions “empirical” and “normative” expectations, she only means that the former is a belief about others’ behaviour, while the latter is a belief about others’ beliefs about one’s behavior.

9

work in either case: the expectation that others think one will

conform does not in and of itself make the expectation

legitimate: it is just a majority opinion about what will

likely be the case. Similarly, a normative belief (that others

think one ought to conform) will not be viewed as reasonable

unless one thinks that the majority opinion itself is a source

of legitimacy. In both cases, legitimacy or reasonableness in

Bicchieri’s sense seems to be ultimately grounded in the

majority opinion and the majority opinion alone can motivate

only if one already has a conformist preference of types (i)

and (ii). For this reason, her distinction between (i) and (ii)

as a generic conformist desire and (iii) as reason, as well as

her emphasis on the latter, does not hold.6

In contrast to Gintis and Bicchieri, Robert Sugden (2000)

models social norms explicitly based on the non-normative

conformist desire. He introduces a resentment hypothesis, according

to which we are motivated to meet the expectations of others

because we are averse to their resentment -- a resentment that

we expect will be provoked by frustrating their expectations.

Sugden takes this to be an empirically confirmable fact about

human psychology; that is, people resent someone who violates

their well-established empirical expectations, and people are

averse to face such resentment themselves. Supposing we do find

others’ expectations “legitimate”, then this will be explained

by wanting to please them because not doing so will provoke

their resentment. So to use Bicchieri’s distinctions, the type

6 In Bicchieri’s examples of social norm engineering, such as anti-littering and a campaign against college students’ binge drinking, people’s behavior changes, not by changing their belief that a certain behavior is legitimate or not, but by changing the perception of how widespread a certain behaviour is. Such an intervention also presupposes a sort of generic preference for conformity.

10

(iii) motivation (to do something legitimate) is grounded in

type (ii) motivation (to please others’ expectations), which in

turn derives from more basic type (i) motivation (to avoid

others’ resentment).

We agree that the aversion toward being a target of resentment

might motivate the conditional preferences to meet others’

expectations in Bicchieri’s model.7 Still, it is questionable

whether aversion of resentment is the type of psychological

feature that can explain a wide range of prosocial behaviour,

including cooperation in one-shot PG games where one has no

reason to expect cooperation as a norm. Moreover it raises the

question of why other resentment should matter to us anyway --

especially if they are strangers. All the same, we do not wish

to argue against the resentment hypothesis at this stage. We

believe that the resentment hypothesis has a legitimate place

as a proximate account, but only once one finds an appropriate

mechanism that explains why resentment should matter to us in

the first place - a task we will turn to in the succeeding

sections of this paper. In section 3 we will see how Sugden

himself suggests what might possibly scaffold such resentment,

but in section 4 and onward we will argue that the social

motivation hypothesis can provide a superior and more basic

account of the resentment hypothesis.

2.2 Team Reasoning Approach

A prominent contender of the social norm approach to explaining

prosocial behavior is the team reasoning approach (Sugden 1993,

7 Sugden (2000) sketches his own model of social norms, but to our knowledge no one, including himself, has tested this model.

11

2000, Bacharach 2006, Coleman et al. 2008). Here the prosocial

behavior is modelled by an individual identifying herself as a member of

a team, and thus playing her part in order to achieve the group’s objective. If a

player identifies herself as a member of a group consisting of

all or most of the players, then it is rational action to

cooperate or coordinate with others, since it will uniquely

maximize the group’s collective payoff. The account then

involves an agency transformation that is distinct from the payoff

transformation of the social norm approach.8

The team reasoning approach needs to specify the epistemic

conditions for the presumed agency transformation -- or shift

from I-mode to we-mode, to use Tuomela’s (2007) terminology.

Bacharach (2006), who offers one of the most systematic

presentations of the team reasoning approach to date, proposes

the ‘interdependence hypothesis’ that roughly states that the

agency transformation is triggered if the agents perceive

common goals that can only be achieved together (for conditions

for agency transformation see Bacharach, 2006, pp. 83-85;

Smerilli 2012). The epistemic conditions of the interdependent

hypothesis are thus provided by such features representing the

uncertain prospect of mutual benefit.

8 Often the team reasoning model is misunderstood as just another other-regarding preference model where “other” regards the group to which an individual belongs. An analysis of the Hi-Lo game shows this is not so. AHi-Lo game is a two-person, two-by-two, normal form, symmetric pure coordination game in which one Nash equilibrium (Hi, Hi), say (2, 2) pareto-dominates the other (Lo, Lo), say (1, 1). The individuals’ payoff transformation that takes into account the group’s payoff changes the payoff of (Hi, Hi) as (4, 4), (Lo, Lo) as (2, 2), which leaves both as Nash equilibria. In contrast, the agency transformation makes (Hi, Hi) asa unique equilibrium of the game.

12

For Bacharach, the shift in framing occurs spontaneously and is

not a matter of choice. Nevertheless, even supposing that

conscious choice typically does not enter into the agency

transformation, the account still seems to require some

proximate account of the psychological features in the

individual agent that are responsible for such a shift

(Pacherie 2011). For unlike a purely perspectival shift between

alternative ways of looking at a gestalt figure, the framing

effect relevant for agency transformation does seem to require

some sustained proximate account with both cognitive and

motivational elements. Consider for example the analogous

reference-dependent framing effect (Tversky and Kahneman 1981),

which is purportedly proximally explained by loss-aversion.

While Bacharach (2006, ch 3) suggests that group selection can

account for why propensities to team reason have arisen under

conditions consistent with the interdependence hypothesis, he

does not give any detail as to what psychological propensities

account for the identification of such features. Indeed, it is

not only team-reasoning accounts who are typically silent on

what motivates the initial and continued attachment to the

group. This problem is common with an influential philosophical

theory of collective intentionality (Gilbert 2013), which

presupposes a joint commitment of the members to their group’s

goals, but does not give an account of proximate motivation for

such commitment (Pacherie 2011).

It is notable that the resentment hypothesis that was intended

as a complement to the social norm approach and also played

some role there is not applicable in the case of team

reasoning. Since a key epistemic condition spelled out in the

13

interdependent hypothesis is a perception of a mutually

beneficial but uncertain outcome, this seems to block the

possibility that it is the fear of resentment that motivates

agent transformation (since why would such circumstances

provoke resentment?). Moreover, presumably collective

intentionality theorists would also hesitate to posit fear of

resentment as a motivational basis of a joint commitment, since

that would reduce the model of collective intentionality to one

of individual norm-following like Bicchieri’s.

2.3 Summary

In this section we have discussed the proximate hypotheses of

the most psychologically informed economic approaches to

prosocial behavior: the social norm and team reasoning

approaches. Our first result is that Bicchieri’s model of

social norms presupposes some motivational account, and that

Sugden's resentment hypothesis is currently the best candidate.

However we also suggested that the hypothesis seems to call for

an explanation of why the resentment of others should matter to

us and noted that it is ill fit to deal with the team reasoning

approach (which lacks any explicit proximate account of its

own). We now move on to some general proximate accounts that

are more promising, but will have reasons to return some of the

hypotheses of this past section (like the resentment

hypothesis) in section 5 to show how they can be scaffolded on

more basic social motivations.

3. The “fellow-feeling” hypothesis

14

In his 2002 paper, Sugden admits that team reasoning lacks a

motivational dimension just like other rational choice models.

His remedy draws heavily on Adam Smith’s work Theory of Moral

Sentiments (1759) and suggests that Smith’s account of fellow-feeling

“may be able to fill these gaps by explaining the affective

qualities of team thinking.” (2002, p. 82) While Sugden clearly

takes the approach as a complement to the team reasoning

account, rather than the social norm account, we will consider

the fellow-feeling hypothesis as a potential psychological

account of prosocial motivation in general, i.e. for both norm

following and team reasoning.

According to Sugden, Adam Smith suggests that simply doing

things together with others rather than alone creates added

value to activities by giving a certain satisfaction or reward.

This reward is due to participants’ positive affective

awareness of their corresponding sentiments. This is also the

definition of fellow-feeling: the pleasant awareness of corresponding

sentiments or emotions. Smith claims that this fellow-feeling is

rewarding irrespective of the valence of original sentiments.

Thus, fellow-feeling cannot be reduced to mere reflected

emotions: “this correspondence of sentiments of others with our

own appears to be a cause of pleasure, and want of it a cause

of pain, which cannot be accounted for ‘by a theory of

reflected feelings’” (Sugden 2002, p. 72). The agreement of feelings

be they joy or grief, is thus a source of reward or pleasure in

its own right such that fellow-feeling “enlivens joy and

alleviates grief.” (Smith, 2002, 18).

According to Smith, our desire to experience fellow-feelings

motivates us to align our actions and emotions with those of

15

others. Moreover, those that have a history of repeated

interactions will be more likely to experience such fellow-

feelings:

On Smith's account, it is a fact of human psychology that people who

repeatedly interact with one another tend to develop and express common

sentiments. Such common sentiments tend to eventually become the objects

of common approval within the group of interacting people. Thus, the

observed failure of any one member of a social group to uphold the attitudes

of that group will cause pain or unease to other members (this is just the

negative equivalent of the pleasure of mutual sympathy); and it will be

disapproved of. (Sugden 2002, 82-83)

The account can thus be seen as a proximate complement to the

existing prosocial models since people become motivated to

adjust their emotions and behaviour to fit with the communal

“norms of propriety” or else they will forgo opportunities for

fellow-feelings. That is, the pleasure derived from fellow-

feelings motivates individuals to uphold certain shared

emotions and behavioral regularities that are associated with

them. The pleasure associated with fellow-feelings also

motivates individuals to identify with groups that they have

interacted with in the past and to uphold the communal norms of

these groups. Alternatively, as Bicchieri would put it, fellow

-feelings give individuals a reason to comply with empirical and

normative expectations. While a generic desire to meet other

people’s expectations cannot explain our context- and group-

specific prosocial behavior (Bicchieri 2006, 25), the fellow-

feeling hypothesis implies that we are only motivated by those

expectations that have arisen during those previous social

interactions that included shared sentiments. True fellow-

16

feeling is a generic affect for both Sugden and Smith, but

repeated interactions with fellow-feelings are needed for the

generic emotion to reliably motivate people to follow context-

and group-specific social norms.

The affective rewards that the individuals derive from the

interaction and the motivations that precede them, are thus

valuable in their own right. In Sugden’s own words, “we directly

desire to participate in society, and not merely the ends that

it produces” (2002, 85). Nevertheless the motivational power of

fellow-feeling proposal relies on rewards that are only

attainable from shared emotions in social interaction and

dependent on awareness of other participants’ corresponding

emotions. In the next section we will argue that because of

this the fellow-feeling approach is still too restrictive and

that correspondence or mutuality of sentiment might not be a

prerequisite for all experiences of intrinsic social rewards.

The model we present will thus be a weaker and also a more

basic model of social motivation. The social motivation here

derives from social affiliative stimuli that neither demand

aligned emotions.

4. The social motivation hypothesis

The pleasurable reward in the fellow-feeling account was

connected to a distinctive type of social encounter. In such

encounters, emotions of the participants are aligned such that

all may reap enjoyment from it; but why suppose that it is only

on these occasions that sociality is rewarding? There seems to

be a broader array of affiliative stimuli, i.e., interactions

with, or observations of, other humans, which have the

17

potential to generate intrinsically rewarding experiences.

These occasions are often consciously, recorded as warmth,

affection, and other positive affect connected to affiliation.9

Neuroscience also suggests that the set of basic biological and

psychological mechanisms that generate these rewards are

elicited by the same modalities of sensory affiliative stimuli

(i.e. touch), even if the social activities that generate the

sensory stimulation are different (Crews 1998; Mason & Mendoza

1998; Depue & Morrone-Strupinsky 2005). Although activities

involving shared emotional experiences certainly induce these

rewards, several other activities are also capable of this

effect, e.g. those that are more purely coordinative,

explorative, and even more competitive in character (think of

how academics enjoy a verbal sparr even if this activity often

involves antagonistic emotions between the disputants).10

Sometimes the reward may even be elicited by something as

fleeting as someone smiling at you.

It is commonly thought that the role of the neurobiological

substrates underlying the affective rewards of affiliative

stimuli (oxytocin, vasopressin, endogenous opioids) is to

orient and bias attention to very basic forms of mammalian

9 In the theory of interpersonal behaviour, affiliation is one of the two major independent traits, characterized by warmth, positive emotions, andagreeableness (the other is agency characterized by assertiveness and activity). To represent social stimuli that are affiliative, experiments typically use pictures of smiling babies and families. (See Depue and Morrone-Strupinsky, 2005, 315).10 The academics enjoy the sparr and thus get reward from it but their enjoyment is not shared in a robust sense (see Salmela 2012). Their emotion is shared only in a weak sense of being the same type (enjoyment)and having the same intentional object (the sparr). Even if the participants enjoy the sparr, their concerns are directly opposite withinthis activity: both strive to win, but since sparr is a zero-sum game where one agent’s win is another’s loss, their emotions do not correspondwithin the activity. Accordingly, no fellow-feelings emerge.

18

affiliation (parent and child) as well as forming social

relationships more generally (Nelson and Panksepp 1998; Depue

and Morrone-Strupinsky 2005). The experience of rewards

attached to social interaction and affiliation generates social

motivation circuitry that not only motivates us to seek the

pleasure in social interactions (affective reward); it also to

work to establish and facilitate social bonds (Godman 2013).

Many social bonds in other primates are also thought to be

mediated or enhanced by rewards elicited during activities such

as play and grooming (Silk et al. 2009). The social rewards

elicited by affiliative stimuli thus provide a foundation for

the capacity for forming social bonds crucial to human

development and flourishing (see e.g Baumeister 1995).

So how precisely does the experience of a reward attached to

social affiliative stimuli lead to a social motivation? The

thriving general field of motivation research backs up the

Humean assumption about how motivations typically grow out from

an anticipation of a reward (Berridge 2003; Wise 2004). In the

case of social affect, the idea would be that affiliative

stimuli initially becomes “incentively salienced” due to the

biased capacity for experiencing such rewards. The initial link

between affiliative stimuli and social reward is then

strengthened in terms of basic associative learning. That is,

because rewards are typically attached to certain affiliative

stimuli, they become increasingly associated in such a way that

they yield (social) motivations in expectation of distinctively

social rewards. Thus, due to the intrinsic rewards afforded by

affiliation and sociality, we become ever more oriented toward

the social world. But, just like in the case of other rewards,

sometimes, wanting eventually occurs without the accompanying

19

conscious, or unconscious, experience of “liking”, thereby

indicating how motivations can be de-coupled from the

associated reward (Berridge and Robinson 2003). (This may

indeed be the case in many economic games where subjects are

naturally conditioned to a particular social reward that is not

in fact elicited).

The social motivation hypothesis then contends that there is a

particular psychological disposition whose role is to orient us

toward affiliative stimuli, which yields social reward (affect)

and enables the formation of social bonds. In fact, the

hypothesis is nicely captured by Coralie Chevallier and

colleagues who defend a social motivation hypothesis in a

rather different context: namely in the case of autism (more on

this in a moment) :

“Social motivation can be described as a set of

psychological dispositions and biological mechanisms

biasing the individual to preferentially orient to the

social world (social orienting), to seek and take

pleasure in social interactions (social reward), and to

work to foster and maintain social bonds (social

maintaining).” (2012, p.231)

It is important to emphasise that the kind of affiliative

stimuli that people seek and find rewarding depends crucially

on their past experiences and learned associations. Some

individuals may actually become motivated to seek antagonistic

social encounters (e.g. winning academic debates or fighting

with partners); perhaps because they have found other rewards

in such activities or even because antagonism is actually a

20

condition of affiliating with others. On the other hand if one

constantly meets rejection and ostracism in trying to affiliate

with particular others, typically one will feel bad and

eventually gives up, or try to find social rewards elsewhere.

In short, social motivations are contingent upon their history

of particular learned associations and although at a baseline

we take some initial affiliative stimuli to be rewarding in

their own right, we are not committed to claiming that all

social stimuli must remain rewarding and sought for.

In recent years there has been several studies in different

domains of developmental psychology and psychiatry that have

suggested that social motivation arises early in development

and is a crucial feature of human psychology. Social

motivations have been suggested to not only mediate and

facilitate bonding but also to explain autism and the pervasive

tendency for social learning (i.e. learning by copying others,

rather than trial-and-error learning). Here we will only

mention a couple of examples of such convergent evidence.

First, it is well-established that infants are born with a

preferential ability to track and fixate stimuli that appear

human-like and evidence indicates that this is mediated by the

neurological centre for affective processing - the amygdala

(Pierce et al 2001; Young et al 2005). In addition, recent

research shows how the amygdala is involved in quickly

detecting salient and diagnostically relevant facial features

by enabling attentional orienting toward the visual periphery

(Gamer et al 2013). These results suggest that we are

preferentially oriented toward one of the most basic form of

21

social affiliative stimuli -- the facial expressions of others

-- due to the associated affective reward.

Recently social motivation has also become an important rival

hypothesis to the standard ‘theory of mind deficit’ approach to

explaining autism (Chevallier et al 2012). The model suggests

that autistic subjects’ difficulties with social life are

better accounted for by diminished social motivation rather

than a failure of ‘theory of mind’, precisely because the

perception of social affiliative stimuli, like facial

expressions, gestures and vocalizations are not accompanied by

any distinctive type of reward. The lack of social motivation

thus impedes social development and can account for many of the

interpersonal deficits associated with the autistic spectrum

disorder (Maestro et al 2002; Chevallier et al 2012).

Finally, social motivations have been suggested to play a

prominent role in human ontogeny in biasing social learning

over individual learning. Cecilia Heyes has for example

suggested that what explains abundant human social learning

strategy is social motivations and other perceptual biases

which privilege information received via social channels over

any individual learning channels (2012, p. 7). Some evidence

for this comes from work on imitation where social motivations

are a strong contender for explaining the pancultural human

propensity for so-called overimitation, i.e. the tendency to

reproduce a high fidelity match of the model's behavior to the

extent that the efficiency of the task is reduced (Nielsen &

Tomaselli 2009). In cases of overimitation we do not reach the

goal in the most effective way possible, but try to do it in

precisely the way it has been modeled to us. For example,

22

children begin to overimitate at the age of two but by no means

cease to do so when they become adults (McGuigan et al 2012).

Social motivations provide a compelling explanation for why

this is so. That is, reproducing the modeled actions appears to

be more important than merely producing the outcomes more

efficiently, precisely because we are socially motivated to

engage in the activity. (Over)imitation is not merely a means

of learning skills or acquiring information about the world; it

is also a means of engaging with others and attaining social

rewards (Uzgiris 1981; Nielsen & Blank 2011). The social

motivation thesis thus predicts that whenever the opportunity

for social rewards are particularly salient part of the target

for learning, how the effect is produced will matter more than

that it is produced in the most effective way possible.

Taken together this evidence from different strands of

psychology also suggests an evolutionary argument for the

emergence of social motivations. Most likely they originated in

their role of facilitating and maintaining social bonds -- a

role which renders them not exclusively human, but among humans

it is likely that they are a particularly strong motivator

(e.g. non-human animals do not tend to over-imitate). Indeed it

may be part of the explanation for why humans are regarded as

the “hyper-social species”. Of course many tie the trait of

hyper-sociality to the emergence of culture, social norms, and

large-scale cooperation (Boyd & Richerson, 2009) and at the

very least once culture enters into the picture, it is highly

likely that culture will co-opt some of the psychological bias

for social reward into new roles. One is their role in

facilitating social learning and the fidelity of social

transmission; another is further biasing individuals to

23

prosocial behavior. Indeed it is natural to suppose that social

motivations are precisely the sort of psychological capacity

that can be harnessed to achieve common, indeed genuinely

altruistic ends that involve overall material sacrifice for the

individual, such as when social motivations help to facilitate

cooperation and cohesion amongst groups.11

5. Social motivations as scaffolders

The social motivation hypothesis is more general than the

fellow-feeling approach in terms of what type of stimuli yield

an affective social reward. For affiliative stimuli to be

rewarding, there is no requirement of fellow-feeling, or for

that matter shared emotions in social activities. In this

section we will consider the relationship between the two

accounts more closely and review some evidence from synchrony

studies that offers support to our more general social

motivation hypothesis over the fellow-feeling account. However

our more important aim is to argue that social motivations

should be thought of as a scaffold for the emergence of fellow-

feelings and, analogously, for other proximal explanations of

prosocial behavior.12

First to the evidence from synchrony studies. The association

between synchronised behavior, on the one hand, and affective

11 While some cooperative endeavours might only be done by groups, or aremore efficiently performed by groups, and so are not genuinely altruisticas they are aligned with self-interest; others might demand foregoing material self-interests. Researchers disagree on which mechanism (self-sacrificing or self-interested prosociality) has more explanatory relevance (see Guala 2012).12 We are very grateful to an anonymous reviewer for suggesting that the concept of scaffolding might be employed to explain their relation.

24

reward, on the other, has been established in several different

studies (e.g. Kirschner & Tomasello, 2010, Reddish et al.,

2013; for a review, see Knoblich et al., 2011). These

affective rewards are associated with different types of

synchrony, such as synchronized motor representations

(Rizzolatti & Craighero, 2004), body postures and gaze patterns

(Shockley et al., 2009), speech patterns (Fowler et al., 2008),

facial expressions (Chartrand & Bargh, 1999), or heart rate

(Vikhoff et al., 2013) of the interactors. A prominent

sociological tradition argues that synchronous activities that

are ritualistic interactions such as plays, games, dancing,

singing, worship and ceremonies, lack any ulterior purpose or

goal and instead have the function of producing affective

rewards for the participants (e.g. Summers-Effler, 2007;

Collins, 2004, Durkheim, 1912). The Smith-Sugden hypothesis

posits that such rewards are produced via shared emotions and

the concomitant fellow-feelings, i.e., pleasant awareness of

shared emotions. This indeed seems to be the case in rituals,

which provide an ample basis for shared emotions that can be

experienced at different stages of the ritual: excitement

during its performance; joy when it succeeds; fear when its

success is threatened; disappointment at failure; anger at

those disturbing the ritual or violating its rules, and so on

(Parkinson et al, 2005; Helm, 2010, Salmela, 2012).

However, there are many other synchronous behaviors, which are

not institutionalised as rituals but nevertheless constitute

basic building blocks for our daily interactions with others.

These tend to be reflected in experiments where the resulting

synchrony is accidental rather than purposefully achieved by

the participants: individuals walk, tap fingers, rock chairs,

25

and so on, either in or out of synchrony, or they rate

interpersonal rapport between figures or sounds that are

presented as in or out of synchrony (e.g. Hove & Risen, 2009;

Miles et al., 2009; Wiltermuth & Heath, 2009). In these cases,

there is little opportunity for feedback and little history of

interaction or reason to suppose the presence of shared

intentionality. Nevertheless, in the studies cited above

individuals who engaged in synchronous behavior reported

greater feelings of interpersonal connectedness, rapport, and

affiliation, and they sometimes cooperated more in a social

dilemma than people in asynchronous conditions. Because

participants also did not report any shared emotions, it is

difficult to make sense of such findings on the fellow-feeling

accounts (or rather, the fellow-feeling approach is at best

silent about the positive effects of such activities). Instead

the social motivation hypothesis seems better equipped in

accounting for the pleasure and reward elicited in more basic

synchronous activities.

Based on this evidence from different synchrony studies, one

might conclude that the fellow-feeling hypothesis and social

motivation hypothesis are not in competition but rather explain

different kinds of synchrony. However, our proposal is that

social motivations have a more important role than simply being

one alternative explanation. Instead, we contend that they can

themselves explain the emergence of fellow-feelings. They do

this by providing a basic platform for how we come to

anticipate rewards of social interactions, which in turn

informs the design and outcome of future interactions. For

example, if individuals experience more basic and contingent

behavioral synchrony as socially rewarding in its own right, we

26

should also expect them to establish forms of social

interaction in which synchrony and rewards are produced

intentionally. This seems to be precisely what happens in rituals.

Although it is very likely that rituals have the capacity to

elicit fellow-feelings through the participants’ shared

emotions at different stages of the ritual, such rewards are

nevertheless initially enabled at least in part by social

motivations and the anticipation of social rewards.

We thus predict that the opportunity for fellow-feelings, not

just in synchronous activities but more generally, is

scaffolded by the basic types of social affiliative rewards

described by the social motivation hypothesis. One way of

understanding social motivations as a scaffold is in terms of

social motivations becoming a condition for more sophisticated

interactions like rituals. However, the key point here is that

in doing so the social motivations also allow for the emergence

of more complex reward structures such as the rewards of

experiencing shared emotions. In fact, this also seems to be

true of the resentment hypothesis, i.e. the desire to meet, and

the aversion to frustrate, others’ expectations (Sugden). To

the extent that these motivations account for prosocial

behavior, it is because they are enabled by and elaborated on

the more rudimental social motivations. In fact, we might even

have a reason to reconsider moral virtues (Gintis) and the

perceived reasonableness of others’ expectations (Bicchieri),

as legitimate proximate accounts for (some) prosocial behavior

if they analogously seem to depend on the experience and

anticipation of social rewards. In short if our affiliations

with other individuals matter, so will their expectations,

interests and moral claims on us. Hence, social motivations

27

explain why moral virtues and others’ expectations matter to us

or why they, as philosophers like to put it, acquire

“motivational force”. In fact it is the strength of the social

motivation hypothesis that it can function as a plausible

scaffold for a range of different more sophisticated proximate

accounts -- without any of them necessarily being reduced to

the anticipation of an affective reward.

Indeed, given that it is supposed that many social situations

in contemporary societies already are governed by some

identifiable moral, social and prudential norms governing

behavior (or at least identifiable behavioural interactive

patterns), it would be surprising if social motivations were not

involved in scaffolding the development of more sophisticated

preferences such as those described by the existing approaches.

In this sense social motivations are always there, so to speak.

Often by making a prosocial outcome of social interaction -- be

it actual, anticipated, remembered, virtual-- at least, prima

facie inherently rewarding for us. That is not to say that social

motivation could not explain prosocial behavior on its own

without some auxiliary motivational account. On the contrary,

we have argued with empirical evidence that in many cases the

motivation to seek social-affiliative rewards is capable of

motivating prosocial behavior, e.g. in situations where the

conditions of other theories of proximate motivation are not

met. Thus, the social motivation account stands on its own in

addition to providing a scaffold for other proximate explanations

of prosocial behavior.

6. Conclusion

28

We began by noting that many existing economic models of

prosocial behavior have been rather silent on the proximate

psychological mechanisms responsible for the target behavior.

We identified the most psychologically plausible accounts and

then based on this discussion and convergent recent evidence

from several strands of psychological and biological research,

we argued for an alternative social motivation hypothesis.

Finally, we suggested that the social motivation account

provides a plausible scaffold for other more sophisticated

proximate motivations that purport to explain prosocial

behavior in more complex social interactions.

Let us conclude by briefly addressing some possible

implications for economic modelling of prosocial behavior. As

we stated at the outset, the empirical and theoretical input

from different fields in psychology does not unequivocally

adjudicate between competing economic models of prosocial

behavior. Nor have we proposed yet another utility function

based on the social motivation hypothesis. Instead, we provided

a basic and general proximate account; namely, social

motivations that anticipate rewards attached to affiliation and

which can scaffold a host of distinct prosocial motivations,

such as moral virtues, aversion to resentment, pleasure of

mutual emotions, and so on.

Although we do not deny that it is possible to compare

different models with different motivations in clear-cut

experiments (this is one of the points of behavioral

experiments, after all), an upshot of our argument is that

there probably will not be a general, single utility function

that encompasses all prosocial behavior. Instead, our account

suggests a form of pluralism that requires different prosocial

29

preferences to be elicited by contextual features such as the

nature of the game or interaction, the past history of

interaction, the strength of affiliation, and the existing

norms or identified teams in a community.

Still, the social motivation hypothesis also suggests some ways

to organize this pluralism about preferences. First, given the

importance of individual learning history and environment more

generally for the precise development of social motivations (as

described in both section 4 & 5), we should expect some

considerable heterogeneity amongst people in terms of their

expectations and preferences in experimental and naturally

occurring games. A satisfactory theory of prosocial behavior

might try to first identify the different types of agents in

their experiments, and then explain how these different types

impact on the interaction, such as sustaining or declining

levels of cooperation (cf. Gächter 2007). Second, given that

social motivations can scaffold a range of distinct prosocial

motivations, we should expect that each agent often has

conflicting preferences that are elicited by variable salient

contextual cues. Thus a general theory should ideally attempt

to identify some general epistemic conditions for the

elicitation of different preferences. For example Bicchieri’s

conditions for payoff transformation and Bacharach’s conditions

for agency transformation do not necessarily have to be seen as

rival hypotheses, but instead they can both be part of a more

general hypothesis which predicts that different conditions can

elicit different preferences of the same individual. We hope

that our proposal has provided some justification for these

lines of enquiry.

30

While what we have said so far is all broadly consistent with

the rational choice tradition in psychology and economics

(although perhaps suspending some of its ambitions of complete

unification), the social motivation hypothesis might also

provide a more basic challenge to the framework, which is the

following: If affiliative rewards are truly rewards on par with

the various material rewards, it may be unhelpful to define or

operationalize rewards solely in terms of monetary payoffs in

games, as for example social preference models do. After all

the research we have discussed suggests that the

neurobiological basis for such rewards is both partly

distinctive from other reward circuits, including material

ones, and arises early in development and is retained

throughout life. The hypothesis also makes good evolutionary

sense given the importance of social bonding for not only

wellbeing, but, indeed, survival. It is unclear if these ends

can consistently be operationalized using monetary rewards.

Thus, the social motivation hypothesis really has two primary

implications for economic modelling of sociality roles. One is

to posit social motivations as a scaffolder that generates

motivations for norm-following and team reasoning, thus

offering a candidate proximate or at least developmental

explanation of prosocial behavior in the standard sense of

representing a choice to forgo immediate material interests.

The second and more provocative suggestion is that in many

social contexts, including perhaps some experimental ones, the

agent is also concerned with anticipated social-affiliative

rewards quite independently of the material payoff

distribution. In future modelling it might therefore be

worthwhile to try to conceptualize and operationalize social

31

rewards instead of just monetary rewards, for example by

employing the kind of basic affiliative stimuli used in the

psychological research we have drawn on.

References

Asch, S. E. 1955. “Opinions and Social Pressure.” Scientific American, 31–35. Bacharach, M. 2006. Beyond Individual Choice: Teams and Frames in Game Theory. edited by N. Gold and R. Sugden. Princeton: Princeton University Press.

Baumeister, Roy F., and Mark R. Leary. 1995. The need to belong: desire for interpersonal attachments as a fundamental human motivation. Psychological bulletin 117(3): 497. Berridge, K. C. 2003. Pleasures of the brain. Brain and cognition, 52(1): 106-128. Berridge, K. C., & Robinson, T. E. 2003. Parsing reward. Trends in neurosciences, 26(9): 507-513. Bicchieri, C. 2006. The Grammar of Society: The Nature and Dynamics of Social Norms. Cambridge: Cambridge University Press. Binmore, Ken. 2006. Why do people cooperate? Politics, Philosophy and Economics 5: 81-96. Boyd, R., and Richerson, P.J. 2009. Culture and the evolution of human cooperation. Philosophical Transactions of the Royal Society B: Biological Sciences 364.1533: 3281-3288.

Camerer, C. F. 2003. Behavioral Game Theory. Princeton: Princeton University Press.

32

Chartrand. T.L. & Bargh, J.A. 1999. The Chameleon Effect: The Perception-Behavior Link and Social Interaction. Journal of Personality and Social Psychology, 76: 893–910.

Chevallier, C., Kohls, G., Troiani, V., Brodkin, E. S., & Schultz, R. T. 2012. The social motivation theory of autism. Trends in cognitive sciences, 16(4): 231-239. Clavien, Christine & Klein, Rebekka. 2010. Eager for Fairness or for Revenge? Psychological Altruism in Economics. Economics and Philosophy 26: 267-290.

Collins, R. 2004. Interaction Ritual Chains. Princeton and Oxford: Princeton University Press.

Coleman, A., B. D. Pulford and J. Rose. 2008. Collective Rationality in Interactive Decisions: Evidence for Team Reasoning, Acta Psychologica, 128: 387-397.

Crews, D. 1998. The evolutionary antecedents to love. Psychoneuroendocrinology, 23: 751–64. Depue, Richard A., and Jeannine V. Morrone-Strupinsky. 2005. "Aneurobehavioral model of affiliative bonding: Implications for conceptualizing a human trait of affiliation." Behavioral and Brain Sciences 28(3): 313-349. Durkheim, E. 2001[1912]. The Elementary Forms of Religious Life. A new translation by C. Cosman. Abridged with an Introduction and Notes by M.S. Cladis. Oxford: Oxford University Press. Fehr, E. and K. M. Schmidt. 1999. A Theory of Fairness, Competition, and Cooperation. The Quarterly Journal of Economics. 114(3): 817-868 Fowler, C.A., Richardson, M.J., Kerry, L.M., & Shockley, K.D. 2008. In A. Fuchs & V. Jirsa (Eds.) Coordination: neural, behavioral and social dynamics, New York: Springer: 261-279.

33

Gamer, M., Schmitz, A. K., Tittgemeyer, M., & Schilbach, L. 2013. The human amygdala drives reflexive orienting towards facial. Current Biology, 23(20), R918. Gilbert, M. 2013. Joint commitment: How we make the social world. New York: Oxford University Press. Gintis, H. 2010. Social Norms as Choreography, Politics, Philosophy, and Economics, 9(3), 251-263. Gintis, H. 2011. Gene-Culture Coevolution and the Nature of Human Sociality, Philosophical Transactions of the Royal Society B, 366, 878-888. Godman, M. “Why we do things together: the social motivation for joint action”. Philosophical Psychology, 26(4), 588-603. Guala, F. 2012. Reciprocity: Weak or strong? What punishment experiments do (and do not) demonstrate. Behavoiral and Brain Sciences35: 1-59. Guala, Francesco, Luigi Mittone and Matteo Ploner. 2013. Group Membership, Team Preferences, and Expectations. Journal of Economic Behavior and Organization 86: 183-190. Gächter, S. and C. Thöni. 2007. Rationality and Commitment in Voluntary Cooperation: Insights from Experimental Economics, inF. Peter and H. B. Schmid (eds.) Rationality and Commitment. pp.175-208. Oxford University Press. Gächter, S. 2007. Conditional Cooperation: Behavioral Regularities from the Lab and the Field and Their Policy Implications, in B. S. Frey and A. Stutzer (eds.) Economics and Psychology: A Promising New Cross-Disciplinary Field. pp. 20-50. MIT Press. Heyes, C. 2012. "What's social about social learning?." Journal ofComparative Psychology 126(2): 193-202.

34

Helm, B. 2010. Love, friendship and the self: Intimacy, identification, and the social nature of persons. Oxford: Oxford University Press. Hove, M.J. & Risen, J.L. 2009. It's All in the Timing: Interpersonal Synchrony Increases Affiliation. Social Cognition, 27(6), 949-60. Kirschner, S. and Tomasello, M. 2010. Joint music making promotes prosocial behavior in 4-year-old children. Evolution and Human Behavior 31: 354-364. Knoblich, G., Butterfill, S., & Sebanz, N. 2011. Psychological Research on Joint Action: Theory and Data. In The Psychology of Learning and Motivation, Academic Press 54: 59-102.

Lisciandra, C., Postma-Nilsenova, M. & Colombo, M. 2013. Conformorality. A Study on Group Conditioning of Normative Judgment. Review of Philosophy and Psychology 4(4):751-764.

Maestro, S., Muratori, F., Cavallaro, M. C., Pei, F., Stern, D., Golse, B., & Palacio-Espasa, F. 2002. Attentional skills during the first 6 months of age in autism spectrum disorder. Journal of the American Academy of Child & Adolescent Psychiatry, 41(10), 1239-1245. Mason, W. & Mendoza, S. 1998. Generic aspects of primate attachments: Parents,offspring and mates. Psychoneuroendocrinology 23:765–78. McGuigan, N., J. Makinson, and A. Whiten. 2011. "From over‐imitation to super‐copying: Adults imitate causally irrelevant aspects of tool use with higher fidelity than young children." British Journal of Psychology 102.1: 1-18. Merrit, J.M., J.M. Doris and Harman, G. (2010) “Character” in J.M. Doris, The Moral Psychology Handbook. Oxford University Press.

Miles. L.K., Nind, L.K. & Macrae, N. 2009. The Rhythm of Rapport: Interpersonal Synchrony and Social Perception. Journal ofExperimental Social Psychology 45(3): 585-9.

35

Nelson, E. E., & Panksepp, J. (1998). Brain substrates of infant–mother attachment: contributions of opioids, oxytocin, and norepinephrine.Neuroscience & Biobehavioral Reviews, 22(3), 437-452. Nielsen, M., and C. Blank. 2011. "Imitation in young children: When who gets copied is more important than what gets copied." Developmental Psychology 47.4: 1050. Nielsen, M., and K. Tomaselli (2010) Overimitation in Kalahari Bushman children and the origins of human cultural cognition, Psychological Science 21(5): 729-736. Pacherie, E. (2011). Framing joint action. Review of Philosophy and Psychology, 2(2), 173-192. Parkinson, B., Fischer, A., & Manstead, A.S.R. 2005. Emotion in social relations: cultural, group, and interpersonal processes. New York: Psychology Press. Paternotte , C. and Grose, J. 2012. Social Norms and Game Theory: Harmony or Discord? British Journal for the Philosophy of Social Science 64(3): 551-87.

Pierce, K., Müller, R. A., Ambrose, J., Allen, G., & Courchesne, E. (2001). Face processing occurs outside the fusiformface area'in autism: evidence from functional MRI. Brain, 124(10), 2059-2073.

Reddish, P., Fischer, R., & Bulbulia, J. 2013. Let’s Dance Together: Synchrony, Shared Intentionality, and Cooperation. PLoS ONE 8(8): e71182. doi:10.1371/journal.pone.0071182

Rizzolatti, G. & Craighero, L. 2004. The Mirror-Neuron System. Annual Review of Neuroscience, 27, 169-192.

36

Salmela, M. 2012. Shared Emotions. Philosophical Explorations. 15(1), 33-46 Schnall, S., J. Benton and S. Harvey. 2008. With a Clean Conscience. Psychological Science 19(12):1219-1222.

Shockley, K.D., Richardson, D.C., & Dale, R. 2009. Conversationand Coordinative Structures. Topics in Cognitive Science, 1(2), 305-319.

Silk, J. B., J. C. Beehner, T. J. Bergman, C. Crockford, A. L. Engh, L. R. Moscovice, R. M. Wittig, R. M. Seyfarth, and D. L. Cheney. 2009 "The benefits of social capital: close social bonds among female baboons enhance offspring survival." Proceedings of the Royal Society B: Biological Sciences 276, no. 1670: 3099-3104. Smerilli, A. 2012. We-thinking and vacillation between frames: filling a gap in Bacharach’s theory. Theory and Decision 73(4):539–560. Smith, A. (2002[1759]) The Theory of Moral Sentiments. Cambridge: Cambridge University Press Sugden, R. 1993. Thinking as a Team: Towards an Explanation of Nonselfish Behavior. Social Philosophy and Policy 10(1):69-89. Sugden, R. 2000. The motivating power of expectations. In Rationality, rules, and structure (pp. 103-129). Springer Netherlands. Sugden, R. 2002. Beyond Sympathy and Empathy: Adam Smith’s Concept of Fellow-feeling. Economics and Philosophy, 18, 63-87. Summers-Effler, E. Ritual Theory. In J.E. Stets & J.H. Turner 2007 (Eds): Handbook of the Sociology of Emotions, pp. 135-54. New York: Springer. Tan, J. H. and Zizzo, D. J. (2008). Groups, cooperation and conflict in games. Journal of Socio-Economics, 37(1):1 – 17.

37

Tversky, Amos & Kahneman, Daniel. 1981. The framing of decisions and the psychology of choice. Science 185: 1124-1131. Tuomela, R. 2007. The philosophy of sociality. New York: Oxford University Press. Uzgiris, I. C. 1981. "Two functions of imitation during infancy." International Journal of Behavioral Development 4 (1): 1-12. van Baaren, R., Janssen, L., Chartrand, T., & Dijksterhuis, A. (2009). Where is Love? The Social Aspects of Mimicry. Philosophical Transactions of the Royal Society B, 364, 2381-2389. Vikhoff B, Malmgren H, Åström R, Nyberg G, Ekström S-R, EngwallM, Snygg J, Nilsson M and Jörnsten R. 2013. Music structure determines heart rate variability of singers. Frontiers in Psychology4:334. Wiltermuth, S.S. & Heath, C. (2009). Synchrony and Cooperation,Psychological Science, 20(1), 1-5. Wise, Roy A. "Dopamine, learning and motivation." Nature Reviews Neuroscience 5.6 (2004): 483-494. Young, A. W., Aggleton, J. P., Hellawell, D. J., Johnson, M., Broks, P., & Hanley, J. R. (1995). Face processing impairments after amygdalotomy. Brain, 118(1), 15-24.

38